DOI QR코드

DOI QR Code

Analysis of Commute Time Embedding Based on Spectral Graph

스펙트럴 그래프 기반 Commute Time 임베딩 특성 분석

  • 한희일 (한국외국어대학교 정보통신공학과)
  • Received : 2013.09.04
  • Accepted : 2013.11.29
  • Published : 2014.01.31

Abstract

In this paper an embedding algorithm based on commute time is implemented by organizing patches according to the graph-based metric, and its performance is analyzed by comparing with the results of principal component analysis embedding. It is usual that the dimensionality reduction be done within some acceptable approximation error. However this paper shows the proposed manifold embedding method generates the intrinsic geometry corresponding to the signal despite severe approximation error, so that it can be applied to the areas such as pattern classification or machine learning.

본 논문에서는 파형 신호와 이미지 등에서 패치를 추출하고 이를 패치 그래프로 구성한 다음, 이로부터 각 패치 간의 컴뮤트 타임을 구하여 이에 기반한 임베딩 기법을 구현하고, 가장 널리 이용되는 PCA(principal component analysis) 임베딩 결과와 비교 분석한다. 임베딩에서 차원을 줄일 경우 원 임베딩과 축소된 차원의 임베딩 간에는 오차가 크지 않도록 차원을 결정하는 것이 일반적이다. 하지만 본 논문에서 구현한 임베딩 방식은 삼차원 이하로 줄여 오차가 80~90%를 상회하여도 축소된 차원의 임베딩 공간에서 각 신호 고유의 기하 구조를 생성하므로 패턴 분류나 기계 학습 등의 응용 목적에 활용 가능함을 실험으로 확인한다.

Keywords

References

  1. H. Hahn, "Proposing a Connection Method for Measuring Differentiation of Tangent Vectors at Shape Manifold," Journal of Korea Multimedia Society, Vol. 16, No. 2, pp. 160-168, 2013. https://doi.org/10.9717/kmms.2013.16.2.160
  2. J.B. Tenenbaum, V. deSilva, and J.C. Langford, "A Global Geometric Framework for Nonlinear Dimensionality Reduction," Science, Vol. 290, pp. 2319-2323, 2000. https://doi.org/10.1126/science.290.5500.2319
  3. T.F. Cox and M.A.A. Cox, Multidimensional Scaling, Chapman and Hall/CRC, 2001, USA.
  4. S.T. Roweis and L.K. Saul, "Nonlinear Dimensionality Reduction by Locally Linear Embedding," Science, Vol. 290, pp. 2323-2326, 2000. https://doi.org/10.1126/science.290.5500.2323
  5. D.L. Donoho and C. Grimes, "Hessian Eigenmaps: New Locally Linear Embedding Techniques for High Dimensional Data," Proc. the National Academy of Sciences, pp. 5591-5596, 2003.
  6. M. Belkin and P. Niyogi, "Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering," Advances in Neural Information Processing Systems, Vol. 14, pp. 585-591, 2001.
  7. M. Belkin and P. Niyogi, "Laplacian Eigenmaps for Dimensionality Reduction and Data Representation," Neural Computation, Vol. 15, No. 6, pp. 1373-1396, 2003. https://doi.org/10.1162/089976603321780317
  8. R.R. Coifman and S. Lafon, "Diffusion Maps," Applied and Computational Harmonic Analysis, Vol. 21, pp. 5-30, 2006. https://doi.org/10.1016/j.acha.2006.04.006
  9. F. Chung, Spectral Graph Theory, American Mathematical Society, 1997, USA.
  10. K.S. Ni and T.Q. Nguyen, "A Model for Image Patch-Based Algorithms," ICIP2008, pp. 2588-2591, 2008.
  11. M. Zontak and M. Irani, "Internal Statistics of a Single Natural Image," CVPR2011, pp. 977-984, 2011.
  12. K.M. Taylor, The Geometry of Signal and Image Patch-Sets, Ph D Thesis of University of Colorado, 2011.
  13. H. Qiu and E.R. Hancock, "Clustering and Embedding using Commute Times," IEEE Trans. PAMI , Vol. 29, No. 11, pp. 1873-1890, 2007. https://doi.org/10.1109/TPAMI.2007.1103

Cited by

  1. Proposing the Methods for Accelerating Computational Time of Large-Scale Commute Time Embedding vol.52, pp.2, 2015, https://doi.org/10.5573/ieie.2015.52.2.162
  2. 파형 신호 공간의 위상 구조 분석 vol.19, pp.2, 2014, https://doi.org/10.9717/kmms.2016.19.2.146